Turl changed the topic of #linux-sunxi to: Allwinner/sunxi development discussion - Don't ask to ask. Just ask and wait! - See http://linux-sunxi.org | https://github.com/linux-sunxi/ | Logs at http://irclog.whitequark.org/linux-sunxi
n01 has quit [Ping timeout: 260 seconds]
m123 has joined #linux-sunxi
<m123> "cuando te encuentres solo(a), mira dentro de ti, hay un corazon, quizas tu mejor amigo y te quiere conversar" bienvenidos: http://castroruben.com *temo_a_un_ser_sin_rival*
m123 has quit [Client Quit]
HdkR has quit [Ping timeout: 252 seconds]
AreaScout has quit []
HdkR has joined #linux-sunxi
Black_Horseman has quit [Ping timeout: 272 seconds]
rellla has quit [Ping timeout: 260 seconds]
hramrach has quit [Ping timeout: 240 seconds]
hramrach has joined #linux-sunxi
nedko has quit [Quit: kernel panic]
<WarheadsSE> iteaduino's have arrived
Quarx has joined #linux-sunxi
Gerwin_J has quit [Quit: Gerwin_J]
Quarx has quit [Read error: Connection reset by peer]
windyan has quit [Ping timeout: 272 seconds]
Gerwin_J has joined #linux-sunxi
egbert has quit [Disconnected by services]
egbert has joined #linux-sunxi
bsdfox has quit [Ping timeout: 260 seconds]
bsdfox has joined #linux-sunxi
kivutar has quit [Read error: Operation timed out]
Gerwin_J has quit [Quit: Gerwin_J]
drachensun_ has quit [Ping timeout: 240 seconds]
bsdfox has quit [Ping timeout: 272 seconds]
drachensun_ has joined #linux-sunxi
bsdfox has joined #linux-sunxi
Gerwin_J has joined #linux-sunxi
_hipboi_ has joined #linux-sunxi
hipboi_ has quit [Ping timeout: 252 seconds]
windyan has joined #linux-sunxi
JohnDoe_71Rus has joined #linux-sunxi
TheSeven has quit [Disconnected by services]
[7] has joined #linux-sunxi
lioka has quit [Ping timeout: 245 seconds]
wolfy has joined #linux-sunxi
bsdfox has quit [Ping timeout: 252 seconds]
bsdfox has joined #linux-sunxi
TomiK has joined #linux-sunxi
TomiK has quit [Client Quit]
Gerwin_J has quit [Quit: Gerwin_J]
wolfy has quit [Ping timeout: 252 seconds]
windyan has quit [Quit: 离开]
bsdfox has quit [Ping timeout: 252 seconds]
bsdfox has joined #linux-sunxi
Black_Horseman has joined #linux-sunxi
Xaros has joined #linux-sunxi
vector80 has quit [Remote host closed the connection]
Black_Horseman has quit [Ping timeout: 260 seconds]
bamvor has quit [Ping timeout: 246 seconds]
bamvor has joined #linux-sunxi
<oliv3r> lo
bamvor has quit [Ping timeout: 272 seconds]
bsdfox has quit [Ping timeout: 272 seconds]
vector80 has joined #linux-sunxi
bamvor has joined #linux-sunxi
<oliv3r> vector80: emmc should be treated as a regular SD card
<oliv3r> but it all depends on where it is connected
<vector80> oh, so you may need to see my sysconfig.fex
<vector80> Mr hendrik mentioned that he need one header from the image to allow my board to be added to uboot-sunxi
bsdfox has joined #linux-sunxi
<vector80> I am trying MMC slot 2 in the config file
<vector80> Lets see
<oliv3r> vector80: if yyou have something installed and it runs right now, you can use a10-meminfo
<oliv3r> to dump those memory setting
<oliv3r> fex file is needed too yeah (both on ML is best)
<oliv3r> anyway
<oliv3r> the a20 boots the following devices in that order
<oliv3r> mmc0, nand, mmc2, spi-NOR
<oliv3r> because of that, its important to know to which pins your emmc is connected
<oliv3r> i would actualyl expect the emmc be connected to mmc2 as you suggest
<oliv3r> allowing mmc0 to be a regular SD card
n01 has joined #linux-sunxi
fredy has quit [Excess Flood]
fredy has joined #linux-sunxi
shineworld has joined #linux-sunxi
n01 has quit [Read error: Connection reset by peer]
<wens> Turl: i will send out v2 of output clock series later
<wens> my system doesn't play nice with accented characters
<oliv3r> wens: i forgot about your patch yesterday; i will try to remember today
<oliv3r> i swear
<wens> so if you're including them in the clock series pull, could you add your acked-by?
<wens> oliv3r: :)
<vector80> oliv3r: Should I send my fex file to where? To you?
<vector80> oliv3r: I have succesfully compiled u-boot and be able to pack with my LinuxBSP, and burn to my board
<vector80> It worked, and when I run "mmcinfo" it dedect my eMMC
<vector80> Also uboot ping is working fine
<vector80> However
<vector80> Here is I got stuck: http://pastebin.com/1CuCFK9B
popolon has joined #linux-sunxi
notmart has joined #linux-sunxi
notmart has joined #linux-sunxi
notmart has quit [Changing host]
Xaros has quit [Ping timeout: 248 seconds]
<oliv3r> vector80: the malinig list thanks
<oliv3r> vector80: has this board ever booted for you?
<oliv3r> i see boot0, boot1 and u-boot starting here
<oliv3r> all seems fine, except that whatever you have told u-boot to boot, isn't formated ext2; anyway, are you using OUR u-boot ontop of THEIR boot loader?
<oliv3r> it might work, interesting concept, but it's probably wrong
<oliv3r> back up whatever you have ont he emmc first ;)
<oliv3r> especially boot0 and boot1
<oliv3r> then dd if=u-boot-sunxi-with-spl.bin of=/dev/emmc bs=1024 seek=8
<oliv3r> now how you get to dd that i'm not sure
<oliv3r> IF you can boot from SD card, it should be relativily easy of course
<oliv3r> do the dd onto your sd card, install a kernel with initramfs
<oliv3r> dd extract bootloader from the booted sd card onto the internal emmc
rellla has joined #linux-sunxi
paulk-aldrin has joined #linux-sunxi
<mnemoc> moin
<oliv3r> mnemoc: maxima + php working? :D
<mnemoc> ok... let me look at that now
<oliv3r> not working today?
<mnemoc> nope, forced vacations :\
<mnemoc> "company shutdown"
<vector80> oliv3r: thank you
<slapin_> mnemoc: are these forced vacations paid?
<vector80> My board have TF card slot too
<mnemoc> yes... like normal vacations
<mnemoc> slapin_: but taken from your vacation days
<vector80> If sunxi devices are booting from TF card first, then how about I try to make one SD card, and try to boot a debian or ubuntu image from the TF card ?
<mnemoc> vector80: note A31 works differently, it needs you to press the recovery button to boot from TF
lioka has joined #linux-sunxi
<vector80> my board is A20
<vector80> my board uses A20
<vector80> Let me find a debian image and try to write to a TF card
<vector80> Lets see what happens
tomboy65 has joined #linux-sunxi
tomboy64 has quit [Ping timeout: 260 seconds]
ZetaNeta has joined #linux-sunxi
kivutar has joined #linux-sunxi
JohnDoe_71Rus has quit [Read error: Connection reset by peer]
Vanfanel has joined #linux-sunxi
JohnDoe_71Rus has joined #linux-sunxi
shineworld has quit [Remote host closed the connection]
tomboy65 has quit [Ping timeout: 240 seconds]
shineworld has joined #linux-sunxi
<oliv3r> mnemoc: ah same here, tomorrow; today is a regular work day
jinzo has joined #linux-sunxi
_hipboi_ has quit [Quit: Leaving]
Vanfanel has quit [Quit: leaving]
bamvor has quit [Ping timeout: 246 seconds]
bamvor has joined #linux-sunxi
massi has joined #linux-sunxi
ZetaNeta has quit [Ping timeout: 252 seconds]
<mnemoc> oliv3r: so the 25 is not national holiday in .nl?
bamvor has quit [Ping timeout: 272 seconds]
shineworld has quit [Remote host closed the connection]
bamvor has joined #linux-sunxi
massi has quit [Quit: Leaving]
bamvor has quit [Ping timeout: 264 seconds]
bamvor has joined #linux-sunxi
geecko has joined #linux-sunxi
<oliv3r> 25th is, 24th not so much
<vector80> oliv3r: Thank you for your recommendation
<vector80> I burn a olimex debian image to one TF card
<vector80> and boot from it
<vector80> it works
<vector80> HDMI, ethernet...etc is ok
<vector80> However, in my board, some GPIO pins are used for different purposes
<vector80> So, I only need to change script.bin file?
<vector80> And also, I think this SD image is for 4Gbyte one, but my SD card is 8Gbyte, how to let it use full size?
n01 has joined #linux-sunxi
<libv> vector80: did you check our wiki
<libv> oliv3r: yeah, that's common
atiti_ has joined #linux-sunxi
atiti has quit [Ping timeout: 264 seconds]
atiti_ has quit [Remote host closed the connection]
ZetaNeta has joined #linux-sunxi
atiti has joined #linux-sunxi
Seppoz has joined #linux-sunxi
Seppoz has quit [Changing host]
Seppoz has joined #linux-sunxi
penghb has joined #linux-sunxi
atiti has quit [Ping timeout: 252 seconds]
n01_ has joined #linux-sunxi
n01 has quit [Ping timeout: 252 seconds]
<vector80> You mean olimex wiki ?
<vector80> Or linux-sunxi wiki?
<mnemoc> is this #olimex?
<vector80> Yes, I tried to read linux-sunxi, it is very helpful, thank you very much
<vector80> As I understand,
<vector80> these SD card images are made by linux hosts and dd tools
<vector80> but LiveSuit images are made by "pack"
jinzo has quit [Ping timeout: 272 seconds]
<vector80> So, for example, if I want to write this SD image to my eMMC, I cann't do that, am I right ?
jinzo has joined #linux-sunxi
jinzo has joined #linux-sunxi
jinzo has quit [Changing host]
<mnemoc> you can not dd them, but you can repartition your nand/eMMC manually, mount the partitions within the dd image, and copy the data
n01_ has quit [Ping timeout: 272 seconds]
n01 has joined #linux-sunxi
<vector80> How to repartition my eMMC manually ? Is that in the wiki ?
<mnemoc> sunxi-tools has a reparitioning tool for allwinner's partition table
<vector80> ok let me find it
<mnemoc> but be aware that the sunxi u-boot doesn't support allwinner's nand partitions. so you have to use the lichee-dev branch
<mnemoc> which is a pitty as allwinner u-boot doesn't pass the memory data properly and so it's not compatible with the open kernels.... unless you fix them
rellla has quit [Ping timeout: 260 seconds]
<mnemoc> and afaik u-boot-sunxi's lichee-dev only supports A10
<libv> nand is a bit of a dark stop still
ZetaNeta has quit [Ping timeout: 272 seconds]
<libv> in our wiki
<mnemoc> oliv3r:
kivutar has quit [Read error: Operation timed out]
<oliv3r> mnemoc: :
<mnemoc> oliv3r: /q
<oliv3r> vector80: you are actually very lucky to have emmc; it makes everything very very simple
<oliv3r> boot from your SD card, copy over the bootloader
<oliv3r> theoretically; done
<oliv3r> vector80: don't worry about sunxi-tools for the partitioning tool; i don't think you need it at all when working with emmc; as emmc is just SD card in a chip
<oliv3r> so ignore all what was said about nand; you don't have nand :)
<mnemoc> oh. emmc :D
<mnemoc> missed that detail
<oliv3r> :p
<oliv3r> yeah but quite significant, as that makes the entire setup so much easier
<oliv3r> actually you could fake a setup like that on the olimex quite easily
<oliv3r> the jumper disables nand
<oliv3r> so you can keep a permanent sd card in the 2nd slot
<oliv3r> which should be virtually identical to the emmc
<vector80> yes
<vector80> thank you very much to both of you
<vector80> In the beginnging
<vector80> I was trying to explain this
<vector80> My board doesn't have a NAND
<vector80> It has eMMC and the structure should be similar to TF card
<vector80> But the question is
<vector80> Normally I use USB-OTG port and LiveSuit windows tool to write those images to my eMMC
<vector80> But when I try to input these debian / ubuntu images to LiveSuit, it says "incompatible" image file
netlynx has joined #linux-sunxi
netlynx has quit [Changing host]
netlynx has joined #linux-sunxi
<vector80> I wonder, how can I make these images compatible with LiveSuit ?
<oliv3r> vector80: do you need to flash 1 device or 1000?
<vector80> For the 1st time, it is about 7400
<vector80> And in future it would be more
<oliv3r> ok well
<oliv3r> even so, no problem
<oliv3r> first, you need ot understand how everything works :p
<vector80> yes :)
<oliv3r> so you'll have to dig into our wiki to get up to speed :)
<vector80> hahahaha
<oliv3r> anyway, it really is very easy
<vector80> Ok, first thing you mentioned above
<oliv3r> checkout sunxi-uboot from git
<vector80> Currently one TF card inserted to my board
<vector80> and it boot from TF card
<oliv3r> build your own u-boot
Quarx has joined #linux-sunxi
<vector80> ok, let me do it one by one
<oliv3r> don't try some random image from some random site :p
<vector80> ok cleart
<vector80> clear
<oliv3r> (olimex is cool)
<vector80> 1st thing is
<vector80> when I compile u-boot
<oliv3r> if you know what you are doing, you can and make it work just fine
<vector80> I tried olinuxino-A20 as the board
<oliv3r> you need a u-boot that boots from sd2 instead of sd0; and I don't think we haev one that does that atm
<vector80> yeah ! How can I change u-boot-sunxi source to accomplish this?
<oliv3r> but first things first, you need a u-boot that can boot from SD card, as you can use that to flash the emmc with that
<vector80> ok
<vector80> Let me start
<oliv3r> maybe you should bring your engineer here to talk about that? :p
<mnemoc> we accept free boards as well :p
<vector80> Bad news is that I am that "engineer" :)
<mnemoc> :)
<oliv3r> yes, send us a free board and we'll do the work for you :p
<vector80> Where are you located ?
<oliv3r> NL
<vector80> cool
<vector80> How can I find your contact info ?
<oliv3r> who designed the board?
<slapin_> ahem
<vector80> Some part is ours
<slapin_> wassup?
<oliv3r> oliver@linux-sunxi.org
<oliv3r> slapin_: hey dud
<vector80> clear
<slapin_> oliv3r: hi!
<vector80> Let me send you one email
<slapin_> I see people talking about free boards
<vector80> But I must tell you something
<oliv3r> slapin_: if you fix mtd; you get one from me :p
ZetaNeta has joined #linux-sunxi
<vector80> I have 2 pcs Olinuxino, 2 pcs Hummingbird, 4 pcs RPi....etc
<oliv3r> hummingbird?
<oliv3r> yeah i have an olinuxino-micro
<vector80> Our board have much more hardware than those hobbyist boards
<oliv3r> a few cubieboards, a few tablets
<slapin_> oliv3r: well, I'm getting serious this NY holiday
<vector80> You don't know hummingbird ?
<oliv3r> slapin_: yay
<oliv3r> vector80: merrii hummingbird
<vector80> right
<slapin_> vector80: please explain what your problem is in simpler way
<slapin_> oliv3r: ah
<vector80> slapin_: My problem is
<vector80> In allwinner linuxBSP,
<vector80> they can create linux images by using pack
<vector80> When they do so, LiveSuit can be used to burn those images to my eMMC on my board directly by using USB-OTG
<oliv3r> you can probably do this much easier without the 'pack'
<slapin_> vector80: do you need Android or someGNU/Linux distro?
<vector80> oliv3r: I wonder how to do it without pack
<oliv3r> vector80: you can have a master SD card, and flash it via that, probably faster as you can do more boards
<vector80> slapin_: yes some of those boards are running android, and some will use debian / ubuntu
<oliv3r> vector80: with emmc its easy, if you insist on usb-otg; then we'll have to write some tools, but not a huge issue I suppose
<oliv3r> vector80: haven't gotten a mail yet
<vector80> oliv3r: Yeah, lets find how to do with a "Master SD card"
<vector80> I couldn't write email, I was talking to you, sorry
<vector80> Let me write asap
<oliv3r> vector80: :p
<vector80> wait
<slapin_> vector80: FEL stuff is quite mature, the only thing you need is to write u-boot script for FEL-bootable u-boot
JohnDoe_71Rus has quit [Quit: Miranda IM! Smaller, Faster, Easier. http://miranda-im.org]
tomboy65 has joined #linux-sunxi
<oliv3r> slapin_: has anybody ever written data to an mmc card using FEL mode?
* slapin_ does have HUGE u-boot script here to burn allwinner boot block on NAND flash with just mw.b commands...
<slapin_> oliv3r: /me did
<vector80> What does FEL-bootable means ? As I know, FEL is boot1 that allows us to enter recovery mode to be able to use LibveSuit ?
<slapin_> oliv3r: via FEL u-boot
<oliv3r> slapin_: share those sfcripts :p i'm sure others want it :)
<slapin_> vector80: FEL is cpu bootrom feature
<oliv3r> vector80: the chip has 2 bootloaders embedded in its BROM (Boot ROM)
<vector80> You mean inside the A20 ?
<oliv3r> the first is what you always use, it tries to read from sd, nand, sd2, spi-nor and boots if it finds a valid image
<oliv3r> yes, inside the SoC
<slapin_> oliv3r: I want to prevent anybody from wanting them, as this is totally sick approach to board bringup
<oliv3r> now, if it fails to boot (or you press the 'fel' button) it falls back into an emergency mode, where the USB-OTG port gets used to upload binaries directly to the memory and execute it
<oliv3r> now, when you flash via FEL-mode, you force the device into this failure mode, and upload binaries that do your bidding :)
<vector80> waww
<oliv3r> vector80: now you have to put each device into FEL mode, and then you can flash via your PC
<vector80> So, you mean
<oliv3r> vector80: if you have a master SD card, you pop it in, power on the device, it'll boot, flash the onboard emmc, blink a led if oyu wish when its' done, you extrat the sd card, and done
<oliv3r> so it's easier imo
<vector80> yeah
<vector80> I accept this
<oliv3r> less manual work
<vector80> yes
<vector80> So, need to find the way
<vector80> How to make such Master Sd card
<oliv3r> i'll reply to your email once i get your e-mail :p
<slapin_> vector80: SD card approach is faster in case of under 10000 boards
<vector80> I agree
<slapin_> vector80: otherwise, direct flash is better, in case of >10000 boards
<oliv3r> if you'd have more then 10k boards, you can hire an engineer to perfect the mtd driver and make everybody happy :p
<slapin_> in other words, me
<oliv3r> in case of 1.000.000 boards, it's probably cheaper to have the factory pre-flash your emmc :p
<oliv3r> slapin_: our mtd driver is quite far, it just needs to get OOB support :)
<vector80> writing long email
<slapin_> if somebody could hire me full time to make mtd driver
<slapin_> oliv3r: funny haha
<slapin_> oliv3r: no
<oliv3r> slapin_: i thought it was quite usable with ubfs allready
<mnemoc> oliv3r: let vector80 hire slapin_ and everyone is happy
<slapin_> oliv3r: it only needs to be compatible with filesystems, thst is
<oliv3r> slapin_: how much manhours do you think it would cost to perfect the mtd driver?
<slapin_> oliv3r: jffs2 and others can live without oob, you know, as some flashes simply can't provide even 64 bytes for user
<oliv3r> mnemoc: but vector80 doesn't even have nand on his boards :p
<mnemoc> doens't matter :p
<vector80> I don't need NFTL
* oliv3r doesn't even know what OOB means, other then out of band; i happily repeat it
rellla has joined #linux-sunxi
<mnemoc> oliv3r: out of band
<mnemoc> err
<oliv3r> or out of bounds :p
* mnemoc should read whole sentences
<slapin_> oliv3r: to perfect - about 160, for working showcase - about 20-40
<oliv3r> anyway, i don't know how that connects to flash :p
<mnemoc> oliv3r: it's somhow like DMA
<oliv3r> slapin_: so how many manhours would it take you to finish the mtd driver as a trusted replacement?
NightShade has quit [Ping timeout: 272 seconds]
<slapin_> oliv3r: I never really estimated the effort in this point of view, as it laregely depends on your expectations. I see and I beleive, it s raw performance can be as fast as allwinner driver, or even faster in some cases, BUT that really depends on how you're going to test it. If tested on raw partitions, MTD approach will be faster, or the same as raw reads of large chunks on Allwinner driver. If there is some performance issues, these can be fix
<slapin_> oliv3r: my goal with this driver is to support SLC flashes with Allwinner, and reliable industrial use cases (read embedded market).
<oliv3r> yeah, i'd imagine that the yaffs2 etc drivers be faster then that actually
<oliv3r> slapin_: so we need a sposnsor to sponsor your work :)
Vanfanel has joined #linux-sunxi
<slapin_> oliv3r: I'll be fine anyway, but if supported, that will be really faster, of course :)
<oliv3r> slapin_: :>
<oliv3r> slapin_: well we can always put a mail out on the ML; sponsor wanted :)
<oliv3r> slapin_: i'll make sure to mention it @ FOSDEM
<Vanfanel> hi, is it possible to use different fb layers for mali_native_window?
<Vanfanel> I need to use a layer different than the one used by the text console
<mnemoc> we should have a (protected) hiring page in the wiki
<mnemoc> to know who from the community is available for hiring and for what kind of work
<slapin_> oliv3r: actually the work now will be probably more about sunxi-current shuffling about new mtd updates re jffs2/ubifs as these are rapidly updated for more recent MLC flashes with large ECCs, as they rely far less on OOB.
<Vanfanel> because, when my EGL/GLES code ends, I usually don't get to the text console again, but instead the last drawn frame
<slapin_> oliv3r: as this will make it easier as testing base
techn_ has quit [Read error: Connection reset by peer]
<oliv3r> slapin_: exciting times, and you need this for your work anytime soon?
<slapin_> oliv3r: this is somewhat a part of the work
techn_ has joined #linux-sunxi
<oliv3r> slapin_: good, so your boss sponsors you kind of :D
<Vanfanel> hmmm.. no sunxi fb experts here?
<Vanfanel> where should I ask, please?
<oliv3r> Vanfanel: usually yeah; i guess not atm :)
<oliv3r> Vanfanel: libv and ssvb are your guys
<Vanfanel> oliv3r: well, ssvb and libv are already reading :P
<oliv3r> Vanfanel: then give it time for them to respond ;)
<Vanfanel> I'm asking since yesterday, maybe it's just a matter of waiting
<Vanfanel> yes, oliv3r, I know, sorry to look so impatient :D
<oliv3r> ssvb ^
<Vanfanel> it's just that other guy and I are "porting" RetroArch to the Cubie properly and only this small detail is missing now
<vector80> oliv3r: Pls check your email
<oliv3r> vector80: allready heard it ding :)
<vector80> :)
<Vanfanel> RetroArch would be the best thing (gaming-wise) coming to the Cubie2. It's an API for lagless emulation and also has Quaque and Doom GLES cores :D
<Vanfanel> just missing the "exit to console" stuff
<oliv3r> vector80: you actually reminded me that I worked for Philip's Set top box unit 10 years ago, looking after firmware updates for all of their eastern european's STB's
<vector80> wawww
<vector80> :)
<vector80> Did you like it :)
<oliv3r> replied :p
paulk-aldrin has quit [Remote host closed the connection]
<oliv3r> full screen IRC lol
<vector80> wonderful isn't it ?
<oliv3r> big board btw, the little one looks far more interestign
<vector80> Both are same
<vector80> Just different camera angle
<oliv3r> ah i see yeah
<oliv3r> the 2nd pic has big board around it
<vector80> That is not a board, it is the plastic case
<vector80> As you see, it is kinda market killer :)
<oliv3r> lol we'll see :p
<vector80> Let me reply email
<oliv3r> haha ok
ganbold__ has joined #linux-sunxi
atiti_ has joined #linux-sunxi
atiti_ has quit [Client Quit]
atiti has joined #linux-sunxi
<vector80> oliv3r: The 1st board on my desk is running debian , and the 2nd one with big plastic case is running android now ( FYI )
<vector80> oliv3r: Do you think the LinuxBSP would be useful for us ?
<oliv3r> the allwinner SDK is horrid and horrible; if you want a proper product, you best stay clear of it ;)
<oliv3r> vector80: anyway, IF you have an image created by the allwinner SDK, be it android or debian, it should have setup your memory somewhat properly
<oliv3r> try to download that and run that on it
<vector80> ok wait
<oliv3r> note, that if you are not using the allwinner bootloader + allwinner generated image, the results will be inaccurate
<oliv3r> vector80: from that information, we can start by adding your board to u-boot
kivutar has joined #linux-sunxi
<vector80> Ok then, best working image is now my android
<vector80> Can I run it on my android ?
<vector80> Or, I should try with their LinuxBSP ?
<oliv3r> whatever came from their BSP
<vector80> ok I will try with the android
<oliv3r> it should detect and configure your memory timings properly
<oliv3r> should be no problem
<vector80> oliv3r: Sorry for delay
<oliv3r> ok, that's the slowest memory timing :p
<oliv3r> but it's 'safe'
<oliv3r> you can always play with memory clocks later if needed
<vector80> Can we overclock it?
<oliv3r> of course, but you'd have to test each board to see how high it will clock
<vector80> Such as 432 or something ?
<oliv3r> 432 is generally safe
<oliv3r> what would the board be called; what's the name of it? how should it be added to u-boot
<vector80> Before it was about 432 I think. We experienced some problems to write LiveSuit images, so we reduced it
<oliv3r> how did you reduce it?
<vector80> I didn't see any difference though
<vector80> sysconfig.fex
<oliv3r> livesuit doesn' tlet you change memory timings does it
<oliv3r> ohh
<oliv3r> really
<vector80> I am confused now
<oliv3r> mnemoc: did you know that? I thought we assumed livesuit actually figures out timings by itself?
<vector80> Because I have set it in the sysconfig as 384
<oliv3r> vector80: oh wait, sysconfig.fex is what is used to create script.bin :)
<vector80> but it show 360 here
<oliv3r> vector80: yeah, because livesuit doesn't look at sysconfig for memory timings :p
<oliv3r> it does memory timing tests and checks how fast it can go
<mnemoc> if the fields are NULL LiveSuit somehow fills the blanks
<aep> is the DSP on the a20 usable?
<oliv3r> aep: DPS on A20?!
<mnemoc> if they have data, the values are preserved
<vector80> Audio Codec can be considered as a DSP :)
<aep> oliv3r: dsp. digital signal processor
<oliv3r> aep: since when does the A20 have that
<vector80> Just joking :)
<aep> dunno. said so on the product website
<oliv3r> vector80: ok so stick with 360, and you can always do testing on faster speeds later
<vector80> clear
<vector80> So, what will be the next step ?
<oliv3r> make me work for free appearantly :(
<aep> oliv3r: actually it says so on wikipedia. http://en.wikipedia.org/wiki/Allwinner_Technology#Chipset_Specs
<oliv3r> i need a board name :p
<mnemoc> linux-sunxi-devel/linux-sunxi-devel-latest.tar.xz -> linux-sunxi-devel-20131224T134436-4a4bf1c.tar.xz .... oh... it's alive!
<oliv3r> aep: the datasheet doesn't even mention it :p
<vector80> Lets say... INTERRA-3
<aep> calls it "cedarx" and says there are no drivers :(
<oliv3r> vector80: ok then that will be your name, pastebin me also your script.bin (in fex format ideally)
<vector80> cedarx is a VPU I think
<aep> oh
<aep> i guess you can use that for generic DSP if you have drivers
<mnemoc> there is cedarv and cedara :p
<aep> i need it for audio anyway
<oliv3r> vector80: it is
<oliv3r> you can't do random dps work if that's what your after
<oliv3r> it's not really a dsp in that sense
<aep> right.
<oliv3r> it's a fixed function dsp
<vector80> oliv3r: http://pastie.org/pastes
<aep> not useful for me then :/
<oliv3r> vector80: that url looks worng
<vector80> oh
<vector80> sorry
<vector80> those are bad
<vector80> Let me try dropbox
<vector80> oliv3r: BTW, pls see our GPIO settings.. Those comment out ones are very important because I am driving panasonix relays with those :)
<oliv3r> not touching those
<oliv3r> ok a few things wrong with your fex file
<vector80> Only gpio_ph25 must be pulled UP, because it is used to allow audio
<oliv3r> first, 'A20 PAD application' :p
<vector80> ahahahhahahaa
<vector80> Sure
<oliv3r> then all that chinese comments? i'm sure they are not relevant for you :p
<vector80> These are from AW
<oliv3r> also 'machine' should have read 'interra-3' at the least :p
<oliv3r> yeah but you are supposed to tailor it to your needs :)
<vector80> I agree
tgaz has quit [Ping timeout: 246 seconds]
<vector80> I didn't find time to modify all these
<oliv3r> :)
<oliv3r> it's kinda important :p
<vector80> I was just back to here 1 week ago, long 15 days in china :(
<oliv3r> right so; what i need to know from you now is, do you have an SD card slot on the board?
<vector80> yes
<vector80> And I can boot debian from it :)
tgaz has joined #linux-sunxi
<vector80> Actually I tried to boot android from it, and it works...
<oliv3r> ok, good, what bus is the SD card connected to? MMC0 i assume?
<oliv3r> and your emmc is connected to mmc2?
<vector80> It was very slow when compared to eMMC
<vector80> Yeah very good question !
<vector80> How can I know which one is MM 0 and which one is MMC 2 ?
<oliv3r> follow the pins :p
<vector80> ok wait
<vector80> Let me see schematic
<oliv3r> though if it boots from SD card, mmc0 is probably the sd card, and eMMC is probably sd2
<vector80> I think so
<oliv3r> ctp_twi_addr = ox38
<oliv3r> that looks wrong
<vector80> FT5506
<vector80> and it is working fine
<mnemoc> o instead of 0
<vector80> oh my god
<oliv3r> 0x38 instead of ox38 :p
<vector80> yes
<vector80> hahahha
<vector80> right
<mnemoc> try fex2bin | bin2fex just for the sake of verifying/sanitizing your data
<mnemoc> fexc -I fex -O fex works too
<vector80> that need to download sunxi-tools and compile statically ?
<mnemoc> `make fexc` is enough
<vector80> ok wait
<vector80> Let me find which SD is zero and which is 2 first
<mnemoc> but it will remove all your comments
<oliv3r> default chinese comments it seems
<oliv3r> vector80: mmc0 = PF02 etc; mmc2 is PC06
<oliv3r> mnemoc: shouldn't hurt to commit a commented fex file to the repo right?
<mnemoc> the fex guide in the wiki is supposed to superseed the comments in the template
bfree_ has joined #linux-sunxi
bamvor_ has joined #linux-sunxi
torbenh3 has joined #linux-sunxi
Nikolas has joined #linux-sunxi
<oliv3r> mnemoc: needs someone to keep it up to date :)
<oliv3r> i'll commit the defexed version
popolon_ has joined #linux-sunxi
honx_ has joined #linux-sunxi
lioka has quit [Ping timeout: 246 seconds]
<mnemoc> oliv3r: who keeps the comments in the .fex updated?
dlan_ has joined #linux-sunxi
<mnemoc> the fex guide is easier to keep up to date
<vector80> eMMC is connected to PC8,PC9,PC10.... pins of A20
<oliv3r> mnemoc: the same person!
<mnemoc> :p
<oliv3r> emmc = mmc2 :p
<oliv3r> as expected
<oliv3r> vector80: how much ram does your board have? 1 Gig?
<vector80> yeah
<vector80> Be back in 10 mins
ZetaNeta has quit [*.net *.split]
penghb has quit [*.net *.split]
bamvor has quit [*.net *.split]
popolon has quit [*.net *.split]
awafaa has quit [*.net *.split]
ojn has quit [*.net *.split]
Turl has quit [*.net *.split]
Montjoie has quit [*.net *.split]
bfree has quit [*.net *.split]
dlan has quit [*.net *.split]
torbenh4 has quit [*.net *.split]
honx has quit [*.net *.split]
Nikolas- has quit [*.net *.split]
<oliv3r> vector80: do you have status leds connected to GPIO's?
<vector80> No we don't have
<oliv3r> you should have!
<vector80> While designing the PCBA, I thought we don't need it, why should we ?
<oliv3r> well a status led on the screen is nice :p
<oliv3r> but, when you going to mass-flash them from a master SD card
<vector80> Olinuxino and hummingbird have
<oliv3r> wouldn't it be nice to be able to flash a led that tis 'done' :p
<vector80> maybe cubie also have
<oliv3r> cubietruck has 4 :)
<vector80> waww
ZetaNeta has joined #linux-sunxi
<oliv3r> you can make a 'diskled' out of it :p
<vector80> If I make one hobby board in future, I will do it
<oliv3r> or, make it blink when it's 'done' flashing :)
ZetaNeta has quit [Changing host]
ZetaNeta has joined #linux-sunxi
Turl has joined #linux-sunxi
awafaa has joined #linux-sunxi
lioka has joined #linux-sunxi
Montjoie has joined #linux-sunxi
ojn has joined #linux-sunxi
<oliv3r> so it can be usefull, even if it's an internal SMD one
<vector80> We can do such things in u-boot source, very nice thing, right :)
<vector80> Actually board have two SMD led on it
<vector80> one is for Power
<oliv3r> and the other one?
<vector80> other is for the proprietary circuit that I mentioned in the email
<oliv3r> not connected via a GPIO? :(
<vector80> no unfortunately
Turl has quit [Changing host]
Turl has joined #linux-sunxi
<vector80> So, let me understand more clearly. You get some DRAM information and system information from me, and you will add an entry to boards.cfg in the uboot-sunxi tree ?
<oliv3r> yes
<oliv3r> doing compile tests right now
<vector80> cool !
<wens> oliv3r: *wink*
<oliv3r> wens: yeha i haven't forgotten about you
<vector80> oliv3r: Did anybody try video streaming via gstreamer on A20 ?
<oliv3r> mnemoc: dl.linux-sunxi.org isn't responding :(
<mnemoc> works for me (tm)
<oliv3r> ssh works
<oliv3r> but i can't open it in my browser
<oliv3r> oh no wait
<oliv3r> i lost my connection earlier
<oliv3r> sec
kivutar has quit [Ping timeout: 252 seconds]
<mnemoc> works on my browser too...
<oliv3r> yeah no was my proxy
<oliv3r> my ssh connection got reset; and thus my proxy was poof
<oliv3r> vector80: dd if=thatfile of=yoursdcard bs=1024 seek=8
<oliv3r> that should yield you a bootable SD card
<oliv3r> if that works; we can move on
<vector80> LEt me try
Vanfanel has quit [Ping timeout: 272 seconds]
<vector80> no need to change name to u-boot.bin ?
<oliv3r> name is irrelevant
<oliv3r> you are writing the content of the file to your sd card
<oliv3r> no filesystem nor names are involved ;)
<vector80> ok
<vector80> Should I make fdisk /dev/sdc and delete all partitions, and format it ?
<oliv3r> no
<oliv3r> you are only replacing the bootloader
<oliv3r> so should work in theory
<vector80> I wrote
<vector80> and sync
<vector80> now extracting
<oliv3r> that's good; then the bootloader works and it's good
kivutar has joined #linux-sunxi
<vector80> wawwww
<oliv3r> right
<oliv3r> now
<vector80> booting
<oliv3r> from the SD card ;)
<vector80> Let me save console log
<vector80> wait
<vector80> booting debian now
<oliv3r> not interesting, but you should have 2 mmc devices in /dev right?
<vector80> boot complete
<oliv3r> /dev/mmcblk0 and /dev/mmcblk1?
penghb has joined #linux-sunxi
<vector80> I couldn't understand your question
<oliv3r> vector80: ls /dev/mmc* or /dev/block/mmc*
<oliv3r> vector80: http://sprunge.us/ is probably easier :p
<vector80> ok I didn't know that one
<vector80> I will try it next time
geecko has quit [Ping timeout: 260 seconds]
<oliv3r> vector80: you can then do ls /dev/mmcblk* | curl -F 'sprunge=<-' http://sprunge.us
<vector80> So, now, I should git clone new tree from uboot-sunxi ?
<oliv3r> IF it is internet enabled
<vector80> yes, EMAC is working in this debian image
<oliv3r> anyway, you did ls /dev/mmcblk0, you need to do mmcblk*, we want to see if you have mmcblk1
<mnemoc> even easier if you make that curl call an alias or ~/bin/ script
<oliv3r> i have a sprunge.sh :)
<oliv3r> so i just do ls | sprunge.sh
<vector80> ok sorry urgent issue
<oliv3r> i think i even dropped the .sh
<vector80> I will be back in 15 mins
<vector80> sorry
<mnemoc> mine is just `sprunge`
<oliv3r> yeah smae
<oliv3r> mnemoc: we dont' have eoma68 schematics do we?
<oliv3r> mnemoc: is the eoma68 using really using mmc3?
<mnemoc> afaik the eoma68-a10/a20 cards are closed hw
<oliv3r> ah, ironic
<mnemoc> and designed by wits
Vanfanel has joined #linux-sunxi
<oliv3r> mnemoc: so lkcl makes it impossible for me to use eoma68 hardware :)
lioka has quit [Changing host]
lioka has joined #linux-sunxi
shineworld has joined #linux-sunxi
<oliv3r> mnemoc: a[12] have 4 sd controllers right? mmc0 - mmc3 right? and we boot from mmc0 and mmc2?
<mnemoc> oliv3r: check his git repo... he probably has some stuff there
<oliv3r> i think u-boot counts starting from 1, isntead of 0
<oliv3r> i know the eoma has a bug with the mmc
<oliv3r> but if it uses mmc3 which is the last mmc port, it can't boot from that
<oliv3r> heh, http://rhombus-tech.net/allwinner/a20/boot/script.fex claims it has 3 mmc ports connected :S
<wens> Turl: patch series posted, might need to check spam folder
<oliv3r> wens: i fetched your tree; force update right?
<wens> oliv3r: i haven't pushed the latest version
<oliv3r> let me know when you've pushed your gmac pathes for u-boot; so i can test them and push them
<oliv3r> first you make me feel bad that i had forgotten about you; then you tell me you haven't even pushed anything yet
<wens> oliv3r: i thought you were going to pick the patches from the ML
<oliv3r> vector80: http://dl.linux-sunxi.org/users/oliver/interra3_emmc.bin and put that as a file! on your sdcards FS
<oliv3r> wens: git pull is easier :p
<mnemoc> merges are ugly
<wens> oliv3r: anyway, i spotted a typo in my signed-off-by
<oliv3r> mnemoc: i tend to drop the merge 'patch' from the tree :p
<oliv3r> mnemoc: because I agree, they are ugly
<oliv3r> mnemoc: actually, i cherry pick patches
<wens> oliv3r: pushed :)
<oliv3r> but even that i find nicer then downloading mail; git am bla bla
<mnemoc> git pull --rebase flattens it
<oliv3r> oh i have git pull aliased to that anyway
<mnemoc> cheater
nove has joined #linux-sunxi
<oliv3r> mnemoc: yeah I am aren't I ;)
<mnemoc> o_O
dlan_ is now known as dlan
<oliv3r> ?
* mnemoc needs another coffee :)
<oliv3r> wens: gmac on cb3 works
Vanfanel has quit [Ping timeout: 240 seconds]
Vanfanel has joined #linux-sunxi
<oliv3r> wens: not working on cb2 :(
<oliv3r> dhcp does work though
<oliv3r> wens: hmm, pretty sure i'm doing the tftp thing right, but it doesn't work
<oliv3r> let me recompile and reflash u-boot
<oliv3r> ah, wrong board :p
<nove> where to get that devmem?
<oliv3r> nove: it's in busybox
<oliv3r> or compile devmem2
<oliv3r> that's googleable
<oliv3r> wens: all works; pushing
<oliv3r> wens: i'll push when vector80 can confirm that his emmc u-boot works; so i can push everything i have in my tree atm
<vector80> oliv3r:
shineworld has quit [Remote host closed the connection]
<vector80> oliv3r: Sorry I had an urgent meeting
<vector80> I will be back in 10 mins
<vector80> but couldn't understand your new bin file
<vector80> You updated something ?
<wens> oliv3r: thanks
shineworld has joined #linux-sunxi
shineworld has quit [Remote host closed the connection]
popolon_ has quit [Quit: Quitte]
fredy has quit [Excess Flood]
fredy has joined #linux-sunxi
paulk-collins has joined #linux-sunxi
<vector80> oliv3r: it seems there are 3 partitions in my SD card
<vector80> Which partition I must copy this new file ?
<nove> can't write to VE sram, do something else than SRAM_CTL0_CFG set to 0x0 must be set?
RichiH is now known as help
help is now known as RichiH
<vector80> oliv3r: Thank you soo much for all your helps again, I copied the file to root partition, but I must leave now, traffic seems terrible as usual, hope to see you tomorrow
<vector80> BTW, thank to all of those people who helped.
vector80 has quit [Remote host closed the connection]
wingrime has joined #linux-sunxi
shineworld has joined #linux-sunxi
shineworld has quit []
<nove> is working, i was looking only at the start that full of zeros, more some address and i found i was looking
notmart has quit [Quit: notmart terminated!]
geecko has joined #linux-sunxi
wolfy has joined #linux-sunxi
Quarx has quit [Read error: Connection reset by peer]
<ssvb> nove: got VE SRAM working?
<nove> ssvb, reading from it, but not writing
Gerwin_J has joined #linux-sunxi
<ssvb> nove: ok
<nove> was verifying where in the jpeg encoder the quantization table were been writing to, it has the same offset and fifo data register that the decoder has
Gerwin_J has quit [Client Quit]
Gerwin_J has joined #linux-sunxi
kivutar has quit [Ping timeout: 240 seconds]
<wingrime> oliv3r: ping
<wingrime> nove: pong
paulk-collins has quit [Quit: Ex-Chat]
<nove> wingrime, hi, i have working the jpeg encoder binary, is very simple,, same sub-engine as h264 encoder
<ssvb> wingrime: has anyone seen jemk lately?
Gerwin_J has quit [Quit: Gerwin_J]
<nove> is likely taking a break
tomboy65 has quit [Ping timeout: 248 seconds]
ZetaNeta has quit [Ping timeout: 240 seconds]
wolfy has quit [Ping timeout: 252 seconds]
<wingrime> ssvb: nope
<wingrime> nove: nice
<oliv3r> wingrime: pong
<wingrime> oliv3r: whats with mmc?
<wingrime> oliv3r: also with a23
<oliv3r> wingrime: ah you need an update?
<oliv3r> well we know the BROM boots mmc0, nand, mmc2 and SPI-NOR right?
<oliv3r> mmc1 and mmc3 are unbootable (to the brom anyway, u-boot can boot whatever it wants if you manage to load the SPL)
<wingrime> oliv3r: yes, I busy at work, so with current timezone I can't be there in such time (1 am now)
<wingrime> not so much time in home
<oliv3r> wingrime: yeah i'm missing you around ;)
<oliv3r> wingrime: anyway, it looks like the EOMA68 is using mmc3, which as far as I can see, isn't bootable
<oliv3r> hence my questions in arm-netbook
<oliv3r> then, we have A23, libv has got one (by accident) in a tablet
<oliv3r> it looks right now that it is like sun6i but with mali
<oliv3r> so it's an sun6 + sun7i hybrid
<wingrime> oliv3r: freq?
<mnemoc> 1.8GHz
<wingrime> graphics?
<wingrime> IP?
<wingrime> and still slow?
<wingrime> nice new techprocess
<oliv3r> wingrime: should be faster, memory runs at 552 MHz
<oliv3r> and mali mp2 i think
<oliv3r> this time it actually runs at 1.6 ghz
<wingrime> oliv3r: still no fex, no uart?
Gerwin_J has joined #linux-sunxi
<oliv3r> wingrime: we have fex, uart, root
Vanfanel has quit [Quit: Lost terminal]
atiti_ has joined #linux-sunxi
<wingrime> a20 uboot should start...
<oliv3r> nope
<oliv3r> diff memory controller :)
<wingrime> oliv3r: whats with current mainline mmc?
<oliv3r> we expect it to be sun6i mem controller
<oliv3r> wingrime: hansg is working on it
<oliv3r> wingrime: so should hit soon
<oliv3r> wingrime: your IR driver can start reviews too imo
<wingrime> oliv3r: clocks are ready?
atiti_ has quit [Client Quit]
<oliv3r> wingrime: yeah, they have been submitted
<oliv3r> wingrime: and we have sunxi-devel kernel now, where all 'to go upstream' or being tested patches are living
<wingrime> natural step forward)
<wingrime> so graphics are relay problem
<wingrime> *realy
<oliv3r> i think hansg said he was going to start playing with kms driver
geecko has quit [Ping timeout: 246 seconds]
<Turl> wens, looked ok on a quick glance on the phone
<Turl> wens, I'll have a proper look tomorrow or so; if all goes right I'll include it on my pull to mike
<oliv3r> irc as root; waddanoob!
<oliv3r> (vector80)
Gerwin_J has quit [Quit: Gerwin_J]
Gerwin_J has joined #linux-sunxi
rz2k has joined #linux-sunxi
wingrime has quit [Ping timeout: 252 seconds]
wolfy has joined #linux-sunxi
nove has quit [Quit: nove]
rellla has quit [Quit: Nettalk6 - www.ntalk.de]
ZetaNeta has joined #linux-sunxi
wolfy has quit [Ping timeout: 252 seconds]
Turl has quit [Read error: Connection reset by peer]
Turl has joined #linux-sunxi
ZetaNeta has quit [Ping timeout: 240 seconds]
tomboy65 has joined #linux-sunxi
NightShade has joined #linux-sunxi
tomboy64 has joined #linux-sunxi
ZetaNeta has joined #linux-sunxi
ZetaNeta has joined #linux-sunxi
ZetaNeta has quit [Changing host]
tomboy65 has quit [Ping timeout: 260 seconds]
kivutar has joined #linux-sunxi
Gerwin_J has quit [Quit: Gerwin_J]
Gerwin_J has joined #linux-sunxi
wolfy has joined #linux-sunxi
atiti has quit [Ping timeout: 252 seconds]
Seppoz has quit [Remote host closed the connection]
atiti has joined #linux-sunxi
ganbold__ has quit [Remote host closed the connection]
rz2k has quit [Read error: Connection reset by peer]
jinzo has quit [Quit: Leaving]
popolon has joined #linux-sunxi
popolon has joined #linux-sunxi
<andhe> with all the positive progress going on it seems to me like http://linux-sunxi.org/Mainlining_Effort sections "work in progress" and "Left to be done" are in need of an update....
n01_ has joined #linux-sunxi
n01 has quit [Read error: Connection reset by peer]
atiti has quit [Ping timeout: 264 seconds]
atiti has joined #linux-sunxi